Job stressors and the innovative work behaviour of STEM teachers: serial multiple mediation role of creative self-efficacy and creative motivation

ABSTRACT

STEM education is crucial in the digital intelligence era, imposing new demands and challenges on teachers. However, there is limited focus on the job stress and innovative work behaviour of STEM teachers, particularly on how job stressors influence their innovative behaviour. This study, from the perspective of school organisational administrators, based on the JDCS Model and social learning theory, surveyed 849 primary and secondary school STEM teachers across 11 provinces in China using scales for Job Stress, Creative Self-Efficacy, Creative Motivation, and Teachers’ Innovative Work Behaviour. Results indicated that job demand positively influences innovative work behaviour directly and indirectly through creative motivation and the serial mediation of creative self-efficacy and creative motivation, contributing 38.4%. Job control also positively influences innovative work behaviour indirectly via creative self-efficacy and the serial mediation of creative self-efficacy and creative motivation, contributing 32.2%. Similarly, job support impacts innovative work behaviour indirectly through creative self-efficacy and its serial mediation with creative motivation, contributing 21.9%. Collectively, job demand, control, and support affect STEM teachers’ job stress. It is recommended that school administrators balance job demand and support, adjust job control, and provide targeted support to enhance teachers’ creative self-efficacy and motivation, fostering innovative work behaviour.

KEYWORDS:

• STEM teachers
• Job stressors
• JDCS model
• Creative self-efficacy
• Creative motivation
• Teachers’ innovative work behaviour

Introduction

In recent years, STEM has become a significant trend or pedagogical approach for the innovative development of future education (Hanif et al., 2019; Othman et al., 2022). It plays an indispensable role in promoting innovation and creativity in the future generation (Leroy & Romero, 2021). The impact of teachers’ creativity and innovative work behaviour (IWB) on the development of student creativity has been extensively studied (Apak et al., 2021; Nisraeni et al., 2020). However, there are insufficient researches focusing on the job stress experienced by STEM teachers, and its influence on their IWB.

The stress faced by STEM teachers in all aspects of their jobs, compared to conventional subject teachers, cannot be ignored (Hong et al., 2021), as they are required to possess a broader range of competencies, including proficiency in humanities, arts, science literacy, and interdisciplinary knowledge. In addition, they must possess the capacity to design, execute, and facilitate student involvement in inquiry-based creative activities (Ejiwale, 2013; Zheng, 2017). However, they also face the fundamental dilemma of inadequate resources and support for innovation (Eisenhart & Weis, 2022). Thus, STEM teachers are subject to multiple stresses of job, time, energy, and job support (JS), as well as constraints of role loads and job expectations.

Extensive researches have consistently demonstrated that job stress significantly impacts teachers’ IWB, and that its negative effects tend to outweigh the positive effects. However, previous studies have mainly concentrated on the sources of job stress (Pogere et al., 2019; Skaalvik & Skaalvik, 2017), the psychological mechanisms of teachers’ job stress (Chitra, 2020; Eksi et al., 2020), and the effects of job stressors on job satisfaction and burnout (Chitra, 2020). Moreover, existing studies that have been conducted on STEM teachers mainly focused on teachers’ pedagogical competencies and knowledge structures (Murphy et al., 2019; Sanders, 2012), professional development and collaboration (Gardner et al., 2019), and pedagogy (Diana et al., 2021). In brief, extensive researches have been conducted to investigate the correlation between stress and creativity, and to explore STEM teachers’ experiences. However, there are a dearth of studies examining the impact of job stress on the IWB of STEM teachers. In this context, it is important to ask the following questions: What are the sources of job stress for STEM teachers? How do these elements of job stress affect the IWB of STEM teachers?

This study aims to address these issues from the perspective of school organizational administrators through a questionnaire survey of primary and secondary school STEM Teachers in 11 provinces and cities in China (Beijing, Guangdong, Hubei, Henan, Hunan, Shandong, Chongqing, and so on). First, it utilizes the JDCS theoretical model to comprehensively analyse the various sources of job stress experienced by STEM teachers. The analysis was conducted for three key dimensions: job demand (JD), job control (JC), and JS. Secondly, in alignment with Bandura’s social learning theory, this study explored how these three job stressors influenced STEM teachers’ IWB through the mediating roles of creative motivation (CM) and creative self-efficacy (CSE). Notably, this study effectively integrates the JDCS model with social learning theory, thereby offering a more comprehensive framework for comprehending the correlation between job stress experienced by STEM teachers and their IWB. Additionally, it elucidates the mechanisms through which this [correlation emerges, while also provides a theoretical foundation and the empirical evidence to support the professional development of STEM teachers.

Literature review and research hypothesis

Relationship between job stressors and IWB of teachers

An increasing body of academic literature has examined the potential correlation between stress and creativity (Bahrami-Vazir et al., 2022; Naseem, 2017). Existing studies indicated that job stress may positively impact individual IWB within a specific range, but can also inhibit individual IWB beyond a certain range. For example, on the one hand it is more likely for workers to have creative ideas and novel solutions under moderate levels of stress or workload (Jaiswal & Arun, 2020; W. Ma & Cheng, 2021), and on the other, situations inducing high stress can lead to a decline in creative performance (Byron et al., 2010). However, it is widely agreed that the negative effects of teachers’ job stress often outweigh the positive effects. Empirical evidence confirmed the notion that when teachers’ job stress exceeds a certain threshold, it negatively affects their professional well-being, as well as their physical and mental health, teaching efficacy, and IWB (Daniels et al., 2011; Kokkinos, 2007).

What are the underlying factors contributing to teachers’ job-related stress? How do these factors influence teachers’ IWB? Some researchers have specifically analysed teachers’ occupational stressors, in two dimensions: work context and interaction with students (Pogere et al., 2019; Skaalvik & Skaalvik, 2017), while other scholars have asserted that the sources of teachers’ job stress are the combination of school climate, working conditions, time management, and students’ behaviours (Kyriacou & Sutcliffe, 1978). These studies indicate that there are differences in the classification of job stressors among teachers. In addition, several scholars have explored the psychological mechanisms of teachers’ occupational stress, and discussed its effects of on burnout and job satisfaction (Antoniou et al., 2013; Eksi et al., 2020). In summary, work environment that is external to teachers can give rise to varying degrees of job-related stress, which subsequently influences their IWB through individual psychological factors. Although a number of studies are focused on how teachers’ job stressors influence their job satisfaction, burnout, and intention to leave (Antoniou et al., 2013; Wilhelm et al., 2000), studies on teachers’ job stressors are scattered because of differences in research perspectives or fields. These studies also do not take the role of the relationship between different job stressors into account. Moreover, more studies examined the relationship between job stress and creativity. However, most of them were based on individual development, and rarely considered the mechanism between job stressors and teachers’ IWB from the perspective of school organizational administrators. In China, principals have the dual role of leaders and educators, and they are the leader and the first person who are responsible for the professional development of teachers (Wang, 2015). Principals not only provide policy and institutional support, material support, and emotional support for teachers’ professional development, but also positively influence teachers’ CM as well as the continuation and improvement of innovative practice activities (H. L. Ma, 2011), which has a significant impact on STEM teachers’ IWB.

The “Job Demand-Control-Support” model (JDCS model) is a classic model in the field of job stress research (R. A. Karasek & Theorell, 1990). It is based on R.-A. Karasek’s (1979) “Job Demand-Control” model, which incorporates the “Job support” dimension. The JDCS model suggests that job stressors are the result of the combined effect of external environmental factors such as JD, JC and JS (Johnson & Hall, 1988). Job stress affects individual motivation and behaviour generation through internal psychological mediation (Bakker et al., 2003; Daniels et al., 2011). According to the JDCS model (Johnson & Hall, 1988), JD, JC and JS jointly influence individual job stress and IWB. JD not only influences JC and JS, but also exacerbates individual psychological stress while promoting the generation of individual IWB. As JC and JS increase, a person’s job stress may reduce.

Among them, JD refers to the regulations or expectations of stakeholders for individuals to accomplish their work tasks, JC mainly refers to the degree of control individuals have over their work, and JS refers to the support or assistance that individuals feel or receive from the external environment in their lives and work. Some empirical researches have further proven the applicability and validity of the JDCS model in cross-cultural contexts, and joint thematic studies (Johnson & Hall, 1988; Snyder et al., 2008). Daniels et al. (2011) introduced it into the study of individual IWB, and further verified that JD, JC, and JS significantly affect individual IWB. The above studies not only provide a theoretical framework for analysing teachers’ job stressors but also align with social learning theory in exploring the mechanisms of human behavioural generation.

Therefore, based on the JDCS model and social learning theory, this study, identifies JD, JC, and JS as the three main sources of job stress for STEM teachers, from the perspective of school organizational administrators, to further confirm and investigate the mechanisms of their effects on their IWB. JD in this study refers to the stipulations or expectations of school organizational administrators, parents, teaching and research staff, and other relevant stakeholders on the content, process, methodology, and effectiveness of STEM teachers’ teaching, such as fostering students’ innovative abilities, winning awards in competitions to enhance the school’s characteristics and visibility, and exploring new interdisciplinary teaching models; JC refers to the right of STEM teachers to have autonomy and control in curriculum design and implementation, curriculum evaluation, resource environment, and going out for exchange and learning, etc; JS refers to the support or assistance that STEM teachers feel from the regional education administration, school administrators, teacher-researchers, experts, scholars, and peers for their own STEM teaching, such as policy support, resource environment support, professional development support, and so on; IWB refers to the process by which STEM teachers generate, introduce, and apply innovative ideas or problem solutions of originality, novelty, and value in their teaching work and strive to put these new ideas into practice in teaching situations, such as by adopting new teaching concepts, methods, or strategies. Based on this basis, this study proposed the following three hypotheses:

H1:

JD positively predicted IWB of STEM teachers.

H2:

JC positively predicted IWB of STEM teachers.

H3:

JS positively predicted IWB of STEM teachers.

Mediating roles of CM

Previous studies have indicated that CSE might act as a mediator between work environment (stress) and IWB (Jaiswal & Dhar, 2015; Michael et al., 2011), in addition to being a significant predictor of individual IWB and job stressors. CSE is a critical driver of individual creativity (Karwowski, 2016), and refers to individuals’ subjective judgements or beliefs about their ability to perform and achieve innovative results in their work (Tierney & Farmer, 2002). CSE is positively associated with individual IWB. For example, according to Beghetto (2006), students with higher levels of CSE are more likely to have positive beliefs, and be more creative in their learning and practice. Similarly, Akbari et al. (2021) and Bagheri et al. (2022) proposed that CSE positively affects individual IWB. For another thing, ample evidence has also indicated the crucial importance of perceived work environment (stress) in elevating CSE. For instance, further research has demonstrated that the development of CSE and its role in creative performance may differ among employees in diverse task settings (Tierney & Farmer, 2002). At the same time, in some studies, CSE has also been employed as a mediator or moderator. Research has indicated that the relationship between teachers’ IWB and their initiative personality is mediated by their CSE (M. Li et al., 2017). Social learning theory suggests that humans are neither wholly dominated by environmental stimuli nor driven by internal forces, but rather function through a continuous interaction between environmental, psychological, and behavioural factors, which are both independent and causal of each other (Bandura & Walters, 1977). In this light, teachers’ job stress can be viewed as an emotional experience resulting from environmental factors acting on it. This in turn affects the emergence and development of individual behaviours. These studies provide a theoretical framework for exploring how job stressors affect STEM teachers. Based on the above, this study proposed the following three hypotheses:

H4:

CSE mediates the relationship between JD and IWB of STEM teachers.

H5:

CSE mediates the relationship between JC and IMB of STEM teachers.

H6:

CSE mediates the relationship between JS and IMB of STEM teachers.

Mediating roles of CM

Current studies support the notion that CM can serve as a mediator and as a significant predictor of individual IWB and job performance (Agnoli et al., 2018; Zhu et al., 2018). CM is the intrinsic motivation for teachers to actively participate in innovation work, which can motivate teachers to discover problems and seek solutions (Amabile, 1996). In the first instance, some previous studies have proven that motivated employees had a significant effect on IWB. As an illustration, a study conducted by Nasir et al. (2019) confirmed that motivation fosters IWB. In the second place, the external environment affects an individual’s CM, a systematic review conducted by Soleas (2021) showed that environmental factors can contribute to the CM. Another research also showed that job stressors negatively impact teachers’ CM, such as work conflict, lack of resources, and other stressors that reduce motivation levels, negatively impact teachers’ CM (Saleem et al., 2015). Meanwhile, many studies have also widely used CM as the mediator. For example, Amabile (1996) stated that the work environment influences individual IWB mainly through internal motivation. Paramitha and Indarti (2014) examined the mediating effect of intrinsic motivation on the relationship between environmental support and creativity. Based on the above, this study proposed the following three hypotheses:

H7:

CM mediates the relationship between JD and the IWB of STEM teachers.

H8:

CM mediates the relationship between JC and the IWB of STEM teachers.

H9:

CM mediates the relationship between JS and the IWB of STEM teachers.

Serial multiple mediation role of CSE and CM

The above studies have verified that CSE and CM mediate the external environment and individual IWB. Therefore, we shifted our focus to investigating the connection between CSE and CM in this study. In the field of creativity, CSE and CM serve as the two primary motivational mechanisms that influence individual IWB. Notably, when Bandura proposed his theory of self-efficacy, he also noted that it affects motivation levels, and thus changes thought patterns, actions, and emotional arousal (Bandura, 1982). Thus, it can be seen that the relationship between individual CSE, CM, and IWB is not straightforward. CSE and CM not only directly affect teachers’ IWB, but also CSE influences teachers’ IWB through CM. It is clear from the extensive evidence in the literature that self-efficacy triggers intrinsic motivation in individuals (Alhadabi et al., 2019; Fletcher et al., 1992; Locke & Latham, 2002). CSE not only affects individual IWB but also directly impacts an individual’s CM, emotions, and so on (Cherian & Jacob, 2013; Zhou et al., 2023). Based on the above, this study proposed the following three hypotheses:

H10:

CSE and CM have multiple serial mediation effects on the relationship between JD and IWB of STEM teachers.

H11:

CSE and CM have multiple serial mediation effects on the relationship between JC and the IWB of STEM teachers.

H12:

CSE and CM have multiple serial mediation effects on the relationship between JS and IWB of STEM teachers.

Research model of the present study

As described above, previous studies have explored the relationship between the four variables of job stress, CSE, CM, and IWB, respectively. However, the mechanisms through which job stressors affect teachers’ IWB still need to be determined. Therefore, this study is based on the perspective of school organizational administrators. First of all, according to the JDCS model, we assume that JD, JC, and JS are the three main sources of job stress that can directly influence STEM teachers’ IWB. Then, upon the social learning theory, this study explored how job stressors affect STEM teachers’ IWB, through the mediating roles of CSE and CM in three dimensions: environment, psychology, and behaviour. Among them, job stressors correspond to environmental factors in social learning theory, CSE and CM are psychological factors, and IWB is behavioural factor. Last but not least, drawing from the above discussion, the current study suggests that there could be a series of multiple mediating effects of job stressors (JD/JC/JS)→ CSE→ CM→ IWB. However, it needs to be clarified that teachers’ IWB, in turn, affect the work environment, and that the relationships between environment, psychology, and behaviour are not linear, but rather interacts and influences each other. Therefore, the model of this study is constructed as shown in Figure 1.

Figure 1. The hypothesized model.

Method

Participants

The research group selected STEM teachers in China’s educationally developed provinces through local educators, STEM education alliances, and experimental cooperative schools of STEM education to conduct the survey. Random sampling was used to collect data from Chinese STEM teachers through the WeChat and QQ group of STEM teachers, Wenjuanxing (https://www.wjx.cn/) – a widely used online questionnaire platform – and distributed paper questionnaires. This survey was conducted from 12 January 2022 to 8 May 2022. There were 849 valid replies from 1,056 STEM teachers in primary and secondary schools (females: 488, 57.50%; males: 361, 42.50%) from 11 provinces and cities in China. The poll covered teachers from several grades in Chinese primary and secondary schools. Participants who completed the questionnaire in less than 60 seconds, skipped over any questions, and gave contradictory answers to the reversal question were eliminated. Among the participants, 407 (47.94%) are elementary school teachers, 231 (27.21%) are middle school teachers, and 211 (24.85%) are high-school teachers; 98 (11.54%) held senior titles, 262 (30.86%) held intermediate titles and 189 (22.26%) held junior titles.

Instruments

A questionnaire was developed to collect the necessary demographic information of participants (such as gender, teaching grade, and title), and their perceived job stress, CSE, CM, and IWB in teaching. The participants’ perceived levels of these factors were measured using the Job Stress Survey (JSS), adapted from Sanne et al. (2005); Creative Self-efficacy Survey (CSES), originating from Tierney and Farmer (2002); Creative Motivation Survey (CMS), created by Amabile et al. (1994); and the Innovative Work Behaviours Survey (IWBS), developed by M. J. Li (2018). For JSS, CSES, and CMS, two master’s degree students in Educational Technology translated the original English items into Chinese and modified some of the items. A professor in Educational Technology reviewed the items to improve the clarity and content validity of the questionnaire. For the IWBS, the Chinese version from M. J. Li (2018)was directly adopted. All items in these four measures used 5-point Likert scales, with 1 indicating “completely non-compliant” (or “Not all characteristic of me”), and 5 indicating “completely compliant” (or “Entirely characteristic of me”). There were a total of 36 items across these four measures. These items were tested on a small group of STEM teachers who were similar to the samples.

Job stress

The measures of job stress for primary and secondary school STEM teachers were adapted from Sanne et al. (2005) DCQS. This scale is divided into three dimensions – JD, JC, and JS – and contains 14 items. As the original scale was designed for employees working in an organization, the items were modified to be specific to STEM teachers’ job stress in primary and secondary schools. Some examples of JSS are “Our school has excellent infrastructure and resources for STEM teaching” and “Our school has a well-developed policy system, safeguards or incentives for STEM education”. Cronbach’s alpha was 0.956, suggesting good internal consistency. The CFA results indicated excellent structural validity: x2/df = 3.306, RMSEA = 0.052, CFI = 0.991, GFI = 0.974, NFI = 0.987, and IFI = 0.99.

Creative self-efficacy

The 4-item version of the CSES developed by Tierney and Farmer (2002) was adopted to measure participants’ CSE; for example, “I can promote flexible thinking in students and come up with many different ideas” and “I can help many students become more creative”. Cronbach’s alpha was 0.798, suggesting good internal consistency. The CFA results indicated excellent structural validity: x2/df = 2.724, RMSEA = 0.045, CFI = 0.997, GFI = 0.998, NFI = 0.998 and IFI = 0.997.

Creative motivation

The measures of CM, for primary and secondary school STEM teachers, were adapted from the Work Preference Inventory Scale (WPI) developed by Amabile et al. (1994). This scale is divided into two dimensions – intrinsic and extrinsic motivation – and contains six items; for example, “I like to try my hand at making something new; it gives me a sense of accomplishment or satisfaction” and “Creativity will be more valuable in the future, so I will work on fostering creativity in my teaching”. Cronbach’s alpha was 0.926, suggesting good internal consistency. The CFA results indicated excellent structural validity: x2/df = 3.764, RMSEA = 0.051, CFI = 0.940, GFI = 0.942, NFI = 0.938, and IFI = 0.940.

Innovative work behavior

The IWB measures for primary and secondary school STEM teachers were applied using the scale developed by M. J. Li (2018). This scale is divided into two dimensions, creative generation and creative execution, and contains 12 items. Some examples of IWBS are “I can get students to be more imaginative and develop their creative thinking better” and “I will often take a project-based approach to design and organize teaching and learning activities”. Cronbach’s alpha was 0.956, suggesting good internal consistency. The CFA results indicated excellent structural validity: x2/df = 1.779, RMSEA = 0.030, CFI = 0.997, GFI = 0.988, NFI = 0.993, and IFI = 0.997.

Data analysis

SPSS 25.0, together with AMOS 24.0, were used to analyse the data. Initially, a confirmatory factor analysis (CFA) was used to evaluate the model fit and validity of the latent variables. Secondly, the internal consistency of each scale was ascertained by computing Cronbach’s alpha (α). Third, common method deviation, descriptive statistics, and correlation analysis were conducted to draw an overall picture of participating STEM teachers’ job stress, CSE, CM, and IWB. Finally, the AMOS software Bootstrap technique was used to investigate the mediating effect in more detail.

Results

Descriptive statistics and correlations among the variables

The means of JD, JC, JS, CM, CSE, and IWB were calculated to indicate the teachers’ general status and the means of them were 4.07, 3.89, 3.75, 3.86, 3.96, 3.88, respectively (see Table 1). The absolute values of Pearson product-moment correlation coefficients among these variables ranged from .69 to .86, indicating that significantly strong (r > .50) correlations existed among all six variables.

Table 1. Means, standard deviations, and correlation coefficients.

Correlation coefficients
	M	SD	JD	JC	JS	CM	CSE
JD	4.07	.76
JC	3.89	.83	.81**
JS	3.75	.92	.69**	.79**
CM	3.86	.81	.73**	.77**	.75**
CSE	3.96	.83	.79**	.83**	.76**	.84**
IWB	3.88	.77	.79**	.80**	.74**	.82**	.86**

N = 849.

JD: job demand, JC: job control, JS: job support, CM: creative motivation, CSE: creative self-efficacy, IWB: Innovative work behaviours.

**p<.01.

Assessment of the model fit

The following six indices were used to assess the goodness of fit of the measurement model and the hypothesized research model: chi-square divided by the value of the degree of freedom (CMIN/df), the goodness of fit index [GFI], comparative fit index [CFI], normed fit index [NFI], incremental fit index [IFI] and root mean square error of approximation [RMSEA]. As shown in Table 2, the results indicated an acceptable fitness for both models following the criteria proposed by Fornell and Larcker (1981).

Table 2. Goodness of fit indices for measure and research models.

Model	CMIN/df	GFI	CFI	NFI	IFI	RMSEA
Measure model	3.91	.94	.98	0.97	0.98	.06
Research mode	3.91	.94	.97	0.97	0.98	.06
Recommended value	<5.0	>.90	>.90	>.90	>.90	<.08

Hypotheses testing

The relationships between JD, JC, JS, CM, CSE, and IWB were tested. The parameter estimates of the adjusted research model are presented in Figure 2.

Figure 2. The verified research models. Note. Dashed lines indicate no significant effects.

Figure 2 presents the predicted values of the paths between the variables in the structural equation model and their significance. JD was found to have a positive, significant, and direct effect on IWB (β=.27, p = .026), CM (β=.61, p < .001), CSE (β=.30, p < .001), and JC (β=.94, p < .001), supporting hypothesis H1. JC could negatively and significantly affect CM (β=-.28, p = .014), positively and significantly affect CSE (β=.49, p < .001), but does not positively and significantly predict IWB (β=-.05, p = .607), therefore, H2 was rejected. JS could positively and significantly affect CSE (β=.16, p < .001), but not positively and significantly affect IWB (β=.01, p = .688) and CM (β=-.03, p = .590), thus, H3 was rejected. CSE was found to have a positive and significant effect on IWB (β=.48, p < .001), CM (β=.67, p < .001). CM was found to have a positive and significant effect on IWB (β=.27, p = .018). Hypothesis testing results revealed that two hypotheses (H2 and H3) were rejected, but hypothesis of H1 remained supported (see Table 3).

Table 3. Results of hypothesis tests.

Hypothesis	Hypothesized path	β	S.E.	Result
H1	JD→IWB	0.27	0.026	supported
H2	JC→IWB	−0.05	0.607	rejected
H3	JS→IWB	0.01	0.688	rejected

N = 849.

JD: job demand, JC: job control, JS: job support, IWB: innovative work behaviours.

The coefficient is the standardized coefficient; *p<.05, **p<.01, ***p<.001.

The serial multiple mediation test

The bias-corrected bootstrap method of the AMOS software was applied to further test the serial multiple-mediation roles of CSE and CM between job stressors and teachers’ IWB. The number of iterations was set to 2000, and the effect values were considered significant if the 95% confidence interval of the mean path coefficient did not contain 0 (Wen & Ye, 2014). The results are presented in Tables 4, 5 , and 6.

Table 4. Test results of the serial mediation role of job demand.

Hypothesis	Hypothesized path	estimate	Boot Method SE	95% confidence interval		Result
				Lower	Upper
H4	JD→CSE→IWB	0.258	1.119	−2.900	0.586	rejected
H7	JD→CM→IWB	0.341**	0.169	0.077	0.695	supported
H10	JD→CSE→CM→IWB	0.361**	1.064	0.126	3.401	supported

JD: job demands;, CM: creative motivation, CSE: creative self-efficacy, IWB: innovative work behaviours.

Note: Coefficients are non-standardized; *p<.05, **p<.01, ***p<.001, bootstrap sample size is 2000.

Table 5. Test results of the serial mediation role of job control.

Hypothesis	Hypothesized path	Estimate	Boot Method SE	95% confidence interval		Result
				Lower	Upper
H5	JC→CSE→IWB	0.228*	0.113	0.014	0.444	supported
H8	JC→CM→IWB	0.053	0.054	−0.03	0.181	rejected
H11	JC→CSE→CM→IWB	0.355***	0.102	0.213	0.616	supported

JC: job control, CM: creative motivation, CSE: creative self-efficacy, IWB: innovative work behaviours.

Note: Coefficients are non-standardized; *p<.05, **p<.01, ***p<.001, bootstrap sample size is 2000.

Table 6. Test results of the serial mediation role of job support.

Hypothesis	Hypothesized path	Estimate	Boot Method SE	95% confidence interval		Result
				Lower	Upper
H6	JS→CSE→IWB	0.326***	0.087	0.164	0.501	supported
H9	JS→CM→IWB	0.012	0.025	−0.025	0.077	rejected
H12	JS→CSE→CM→ IWB	0.292***	0.071	0.167	0.444	supported

JS: job support, CM: creative motivation, CSE: creative self-efficacy, IWB: innovative work behaviours.

Note: Coefficients are non-standardized; *p<.05, **p<.01, ***p<.001, bootstrap sample size is 2000.

As Table 4 demonstrates, CM can significantly mediate the relationship between JD and IWB (coefficient = 0.341; 95% bootstrap CI = [0.077, 0.695]; p < 0.01). Thus, H7 was supported. CSE and CM can collectively and significantly mediate the relationship between JD and IWB (coefficient = 0.361; 95% bootstrap CI = [0.126, 3.401]; p < 0.01). Hence, H10 was supported. However, CSE did not significantly mediate the relationship between JD and IWB (coefficient = 0.258; 95% bootstrap CI=[−2.900, 0.586]; p = 0.5). Therefore, H4 was not supported. In summary, JD can affect STEM teachers’ IWB indirectly through CM, but not through CSE. In addition, JD can influence STEM teachers’ IWB through the serial multiple mediating role of CSE and CM.

As Table 5 demonstrates, CM did not significantly mediate the relationship between JC and IWB (coefficient = 0.053; 95% bootstrap CI=[−0.03,0.181]; p = 0.217); therefore, H8 was not supported. CSE can significantly mediate the relationship between JC and IWB (coefficient = 0.228; 95% bootstrap CI = [0.014,0.444]; p < 0.05); Thus, H5 was supported. CSE and CM can collectively and significantly mediate the relationship between JC and IWB (coefficient = 0.355; 95% bootstrap CI = [0.213,0.616]; p < 0.001); Hence, H11 was supported. In conclusion, JC can affect STEM teachers’ IWB through CSE, but not through CM. In addition, JC can also influence STEM teachers’ IWB through the serial multiple mediating role of CSE and CM.

As Table 6 demonstrated, CM did not significantly mediate the relationship between JS and IWB (coefficient = 0.012; 95% bootstrap CI=[−0.025,0.077]; p = 0.545); Therefore, H9 was not supported. CSE significantly mediated the relationship between JS and IWB (coefficient = 0.326; 95% bootstrap CI = [0.164, 0.501]; p < 0.001); Thus, H6 was supported. CSE and CM collectively and significantly mediated the relationship between JS and IWB (coefficient = 0.292; 95% bootstrap CI = [0.167,0.444]; p < 0.001); Consequently, H12 was supported. In summary, JS can have an indirect effect on IWB of STEM teachers through both CSE and CSE acting on CM, but not indirectly through CM acting on the IWB. In summary, JS can affect STEM teachers’ IWB through CSE, but not through CM. In addition, JS can also influence STEM teachers’ IWB through the serial multiple mediating role of CSE and CM.

Discussion

This empirical study suggests that JC, JD, and JS, as the three major sources of STEM teachers’ job stress, can influence teachers’ IWB through the mediating roles of CM and CSE in the context of mainland China. Firstly, based on the JDCS model, this study argued that STEM teachers’ job stressors arise mainly from the combined effects of JC, JD, and JS. Secondly, in accordance with social learning theory, individual behaviours can be affected by both organizational variables and individual psychological factors. Therefore, this study proposed that job stressors do not all directly affect teachers’ IWB, but do so indirectly through the mediating role of CM or CSE. Thirdly, it is essential to note that CM and CSE have serial multiple mediating roles between job stressors and STEM teachers’ IWB.

Effects of JC on IWB of STEM teachers

The findings indicated that although JC cannot directly influence the IWB of STEM teachers, it can positively and significantly influence STEM teachers’ IWB indirectly through the mediating role of CSE. At the same time, it can also positively and significantly influence STEM teachers’ IWB indirectly through the serial multiple mediating role of CSE and CM. For one thing, JC did not directly affect STEM teachers’ IWB. However, it could negatively and significantly affect CM, which is inconsistent with the findings of Unsworth et al. (2005), Daniels et al. (2011), and Cerne et al. (2017). A possible explanation for this might be that some school administrators do not have a deep understanding of the concept or value of STEM education (Kulakoglu & Kondakci, 2023). As Professor Li pointed out, STEM education currently suffers from formalization, over-technicalization and homogenization of values, and the humanistic value of STEM education should be fully emphasized, rather than fixed knowledge or simple mechanical skills (M. Li & Yi, 2022). This may lead teachers to be concerned with “utilitarian” outcomes, limiting their CM or creating “innovation inertia”. In addition, it is noteworthy that both domestic and international scholars have also discovered inconsistent empirical findings concerning JC’s effectiveness in alleviating job stress while testing the JDC model. A research has found that when there is a mismatch between self-efficacy and JC, individuals are unwilling or ill-equipped to utilize JC to cope with the JD, thus increasing job stress (X. Zhao & Zhao, 2010). Another possible explanation might be that JC has a differential meaning for individuals, and excessive JC tends to result in individuals being more inclined to conform to group norms or comply with authority (Fromm, 2014). In China, most public schools are structured and organized using a relatively rigid hierarchical management system, with teachers existing at the end of the overall school structure (Lu, 2020), and school administrators giving teachers more control over their work, leading to lower CM. Nevertheless, within diverse educational contexts and management styles, although an overbearing imposition of job control upon teachers’ tasks may induce “innovation inertia”, a judicious allowance of autonomy to teachers by school administrators, predicated upon the reinforcement of innovative work imperatives, can indirectly exert a constructive influence on teachers’ IWB through CM (M. J. Li et al., 2016). For another, JC can indirectly positively and significantly affect teachers’ IWB through CSE and CM, as Cerne et al. (2017) argued that high JC could help in coping with high JD. In conclusion, although an excessive level of JC may result in “innovation inertia”, a moderate level of JC may positively impact teachers’ IWB indirectly through CSE and CM, while reinforcing the requirements of IWB.

For this reason, school organization administrators should dynamically adjust their JC. Specifically, school organization administrators should learn to dynamically adapt to teachers’ sense of JC. For example, when teachers possess a solid CM, it is imperative to enhance their sense of JC, and provide them with the requisite JS to promote IWB (Yan et al., 2021). Otherwise, school organization administrators should weaken their sense of JC, and provide them with innovative JD or expectations to stimulate innovative external motivation.

Effects of JD on IWB of STEM teachers

The findings suggested that JD can not only significantly and positively influence IWB of STEM teachers but also indirectly through the mediating role of CM. Additionally, JD can positively and significantly influence STEM teachers’ IWB indirectly through the serial multiple mediating role of CSE and CM. On the one hand, JD can directly and significantly positively influence the IWB of STEM teachers, which was in line with the findings of Unsworth et al. (2005), Daniels et al. (2011), Janssen (2004). This result can be explained by the fact that when individuals perceive IWB as essential and valuable, they are more likely to invest time and effort into participating (Unsworth & Clegg, 2010). On the other hand, JD could also indirectly influence the IWB of STEM teachers through CM or CSE acting on CM. Innovative JD usually lead teachers to perceive that the organization expects or requires them to generate innovative ideas related to their work (Unsworth et al., 2005), thus indirectly affecting their IWB. Specifically, when school organization administrators require or encourage teachers to focus on the development of students’ innovative skills through their teaching activities, it not only directly enhances teachers’ CSE and CM, but also contributes to their IWB.

On this basis, school organization administrators should rationalize and optimize their JD. Specifically, the following recommendations are proposed in this study. First of all, school organization administrators should be “far-sighted” about the value and positioning of STEM education, and should have foresight and anticipation in regard to the development and cultivation of STEM (Yang, 2001) to fully support and encourage teachers’ IWB. Moreover, teachers should be provided with a relatively relaxed atmosphere for innovation, which should neither be utilitarian – using rewards as the criterion for assessing teacher performance, nor assigning too much operational tasks to STEM teachers. Finally, administrators should actively encourage teachers to be involved in teaching, training and research programmes related to innovative teaching and learning.

Effects of JS on IWB of STEM teachers

The findings illustrated that although JS could not positively and significantly affect the IWB of STEM teachers, it could positively and significantly affect teachers’ IWB indirectly through the mediating role of CSE. Furthermore, JS can positively and significantly influence STEM teachers’ IWB indirectly through the serial multiple mediating role of CSE and CM, which was consistent with the findings of Cerne et al. (2017) and Tierney and Farmer (2002). There are several possible explanations for these results. A possible explanation for this might be that JS, as the overall awareness of an individual who feels valued, cared for, and supported by the organization, superiors, or colleagues at work (Eisenberger et al., 2001), is one of the sources of motivation for individuals to engage in creative activities (Mumford et al., 2002). Not all JS can influence individual IWB, but rather JS that teachers perceive as beneficial to their own innovation that has a positive impact, as demonstrated by the research of Anggia Paramitha and Indarti (2014). It was noted that this is not the case for the supports from supervisor and family. Another possible explanation for this is that JS has a positive impact on relieving work stress, enhancing CSE, and consequently encouraging CM. For example, organizational resources, such as financial resources, training, and learning, can lead to innovation by augmenting an individual’s CSE (Homberg et al., 2019). Nevertheless, it is worth noting that the total effect of JS on teachers’ IWB was merely 0.106. This suggests that JS, when considered as an isolated external factor, has a relatively limited influence on the IWB of STEM teachers. Thus it is possible to indirectly improve teachers’ IWB by adding various antecedents or mediators (Mustika et al., 2020). Hence, it is imperative to influence STEM teachers’ IWB in conjunction with factors such as JD and JC.

On this basis, school organization administrators should provide JS in accordance with STEM teachers’ individual needs. In particular, the following points were discussed. Above all, school organization administrations must provide appropriate material assistance based on teachers’ individual preferences and specific requirements, and effectively improve material conditions, including the availability of time and necessary equipment, for STEM teachers to actively participate in innovative activities (Yang, 2001). Furthermore, according to the teacher professional development level, it’s necessary to construct the “precision” teacher professional development model of “stratified and phased + teacher innovation practice community”. Last but not least, in consideration of the requirements for teachers’ professional development, administrators could develop an effective incentive and guarantee system, such as the establishment of a highly flexible and adaptable mechanism for the appointment of cadres, labour incentives, and the deployment of teachers, that can be adjusted to suit the prevailing circumstances (H. X. Li, 2000).

The serial multiple mediating role of CSE and CM between Job Stressors and IWB

The results of the study showed support for research hypotheses H5, H6, H7, H10, H11 and H12, suggesting that CSE and CM have a mediating role between job stressors and STEM teachers’ IWB. For example, CSE mediates the link between JC and IWB, suggesting that STEM teachers are more likely to display IWB when they feel in control of their work environment. Similarly, JC, JD and JS positively influence IWB through CSE and CM, which highlights the importance of providing STEM teachers with appropriate challenges and supports. That is, school organization administrators need to provide STEM teachers with policy support, resource support, teacher professional development support, etc., while also giving teachers a certain level of JD, such as fostering students’ innovation and complex problem-solving skills.

In addition, the negative effect of JC on CM (pathcoefficient = −0.28, p < 0.05) may indicate that excessive control limits teachers’ autonomy and thus inhibits their CM. According to research, there is a significant inverted U-shaped curve between JC and individual creativity, where both insufficient and excessive levels of JC can negatively impact individual creativity (Caniëls et al., 2022). Therefore, school organization administrators should have a leading role in controlling the work of STEM teachers (Sheng et al., 2017), rather than prescribing all the detailed demands of their work, and leave STEM teachers with a certain degree of autonomy in their work (Lin & Gao, 2023). According to SDT (Deci & Ryan, 2000; Deci et al., 2017), when the need for autonomy is satisfied, people are intrinsically motivated and in turn become more creative and productive. However, the positive effects of JD on CM (path coefficient = 0.61, p < 0.05) and CSE on CM (path coefficient = 0.67, p < 0.001) reveal that moderate challenge and self-confidence can stimulate teachers’ CM. These findings are in line with G. Zhao et al. (2022) who found that JD and CSE can promote teachers’ IWB by enhancing CM. Finally, JD, JC and JS facilitated the generation of IWB among teachers through the chain mediation of CSE and CM, which are consistent with previous studies (Nan et al., 2016; Tierney & Farmer, 2004).

In summary, CSE and CM play a key mediating role between job stressors and STEM teachers’ IWB. These findings not only support social learning theory, but also enrich the theoretical framework of IWB, as well as providing school organization administrators with practical strategies to promote innovative teaching practices by enhancing teachers’ CSE and CM.

Dynamic modulation of job stressors to promote IWB of STEM Teachers

Admittedly, based on the results of this study’s theoretical model and data analysis, the impact of teachers’ job stressors on STEM teachers’ IWB should be analysed from a holistic, connected, and dynamic perspective. Firstly, school organization administrators should endeavour to balance JD and JS. High JD, if adequately supported, may potentially foster teachers’ IWB (R.-A. Karasek, 1979). Therefore, in addition to heighten JD, school organization administrators also need to offer suitable JS, encompassing professional developmental opportunities and allocating the resources appropriately. Secondly, JC also requires special attention. Too much JC may limit teachers’ innovation space, whereas a moderate level of JC can provide a certain degree of freedom, thus facilitating IWB (Hackman & Oldham, 1976). School organization administrators can flexibly adjust the degree of JC according to teachers’ varying levels of experience and skills. Finally, school organization administrators can provide necessary JS to enhance teachers’ CSE and CM, for example, through incentives, training programmes, or teacher participation in decision-making. In conclusion, school organization administrators should actively facilitate the cultivation of IWB among STEM teachers by flexibly adapting to three key job stressors: JD, JC, and JS.

Limitations and future work

This study explored the influence mechanisms of job stressors on STEM teachers’ IWB from the perspective of school organization administrators, based on the JDCS model in terms of JD, JC, and JS. However, this study has several limitations. Firstly, the environmental and psychological factors affecting STEM teachers’ IWB are not limited to those discussed in this study. For example, changes in demographics (such as gender and age), psychological capital, and organizational traits (such as innovation climate, cohesion, and organizational structure) can also affect the dependent variables. Secondly, this study used self-reports to measure the dependent variable: teachers’ IWB. It did not use more specific methods, such as a mix of quantitative and qualitative research, processes, or formative evaluations, instead of summative evaluations. Thirdly, in terms of the demographic characteristics of the study population, which were chosen as primary and secondary school STEM teachers, it did not examine whether differences in school departments, gender, etc. would affect the results. The differences in the factors influencing STEM teachers’ IWB and the mechanisms of action in out-of-school training institutions were also beyond the scope of this study. This study is based on a survey of STEM teachers in China. Longitudinal studies will be added to future research, such as the mechanisms by which job stressors affect IWB of STEM teachers at different stages of their careers. Qualitative research will also be conducted to understand the real-life experiences of STEM teachers in terms of job stressors, CSE, and IWB through interviews or focus groups with them. Exploring this qualitative data in depth can complement and deepen the understanding of the quantitative analysis. Further studies can also examine the variations in STEM teachers’ job stressors and their influence mechanisms in different cultures or countries.

Implications for practice

This study examines how job stressors affect STEM teachers’ IWB by drawing on the JDCS model and social learning theory. These findings have various implications for teaching practice. First of all, this study integrates the JDCS model with Bandura’s social learning theory, expanding the theory’s potential applications and offering a more comprehensive understanding of the connection between STEM teachers’ IWB and job stressors. Secondly, by exploring the mediating role of CSE and CM, this study describes the relationships between job stressors and IWB, and explains how this relationships arise. Finally, the results of the study can not only provide strong data support for educational policymakers to improve the working environment and conditions of STEM teachers, but also provide new methods and pathways for school organizational administrators to promote the professional development of STEM teachers.

Conclusion

In the context of Industry 4.0 “Preparing for the Future” education, STEM education is defined as a new kind of teaching reform and learning style that benefits students’ 21st-century literacy, including problem-solving, critical thinking, and creativity. Teachers are the “leaders” of students’ creative practices, and this study explores the effect factors and mechanisms of STEM teachers’ IWB, based on the JDCS model and social learning theory. The results of this study indicate that: (1) the rapid development of STEM education has also put forward new requirements and challenges for STEM teachers, who also have to cope with job stress from a combination of external environmental factors, such as JD, JD, and JS; (2) JD, JC, and JS collectively influence STEM teachers’ job stress and can indirectly influence their IWB through the serial multiple-mediation effects of CSE and CM; and (3) School organization administrators can balance STEM teachers’ job stress and promote the generation and development of their IWB by reasonably optimizing JD, dynamically adjusting JC, and providing JS as needed. Overall, this study explored STEM teachers’ creative teaching and related issues, and provides references and lessons for the high-quality development of STEM education and the construction of an innovative teaching force.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This study was supported by the National Youth Project, National Office for Education Science Planning (China). The grant number is [CCA210255].

Notes on contributors

Bin Wu

Bin Wu is a lecturer and master’s supervisor in the field of Educational Technology at Xinyang Normal University. He holds a Ph.D. and his research interests include the basic theories and applications of educational technology, the development of higher-order thinking through technological empowerment, and technology-supported teaching and learning.

Ye Song

Ye Song is a lecturer with a master’s degree at Xinyang Normal University. Her research interests include information-based teaching, maker education, and artificial intelligence education.

Mingzhang Zuo

Mingzhang Zuo is a professor and Ph.D. supervisor in the Faculty of Artificial Intelligence in Education at Central China Normal University. His research interests include the basic theories of educational technology and educational informatization.

Li Zhai

Li Zhai is a master’s student at the School of Educational Science, Xinyang Normal University. Her research interests include the basic theories of educational technology, digital-supported learning innovation, and educational informatization.

Mengsi Zhang

Mengsi Zhang is a master’s student at the School of Educational Science, Xinyang Normal University. Her research interests include the basic theories of educational technology, blended teaching, and the development of students’ critical thinking through technology empowerment.